The definition of extremely large refers to data sets that are too big in terms of volume (too much), and/or velocity (too fast), and/or variety (too many places, too many formats) to be handled using conventional solutions.
In October 2007, data experts gathered at SLAC National Accelerator Lab for the First Workshop on Extremely Large Databases.
As a result, the XLDB research community was formed to meet the rapidly growing demands of the largest data systems.
In addition to the original invitational workshop, an open conference, tutorials, and annual satellite events on different continents were added.
[4] The XLDB organizers started defining a science benchmark for scientific data management systems called SS-DB.